CoGraphNet for enhanced text classification with co-occurrence graph-based representation learning

This paper introduces CoGraphNet, a novel graph neural network architecture for text classification. By constructing word co-occurrence graphs from text corpora, CoGraphNet captures rich relational information between words that traditional sequential models miss.

Key contributions:

• Co-occurrence graph construction strategy for text representation
• Graph-based feature aggregation for document-level classification
• State-of-the-art results on standard text classification benchmarks

Recommended citation: Chen, J., et al. (2025). "CoGraphNet for enhanced text classification with co-occurrence graph-based representation learning." Scientific Reports.